GtDep/ItHadToHappen2 jw May 2009 the great deflation of 1929-33 (almost) had to happen 1

Yüklə 190.65 Kb.
ölçüsü190.65 Kb.
1   2   3

3. Gold and prices, 1914-33

The first eight columns of Table 2 show wholesale prices and commercial bank deposits relative to 1913 for countries for which data are available. Changes in bank deposits may be seen to approximate those for money stocks in the countries for which comparisons are possible. The last three columns of Table 2 show that official gold reserves (in central banks and Treasuries) more than doubled between 1913 and 1928 (10,058/4859 = 2.07). Although less than the nearly three-fold increase in money, even this ratio was overstated.

Table 3. The Gold Stock and its Uses, 1913-39.

(End of year, millions of Troy ounces; % is average annual rate of change from preceding value)

Gold stock

Official reserves


Gold money

Nonmonetary gold











































Sources: Table 2, Kitchen (1929, 1932), Jastram (1977), Edie (1929), U.S. Mint (1940).

Table 3 shows world production and uses of gold between 1913 and 1939. After an average annual rate of 3.2% between 1899 and 1913 (2.9% beginning in the early 1890s), gold production (as a percentage of the stock of gold) fell to 2.2% between 1913 and 1928 (1.9% in the 1920s). The more rapid rise in reserves was made possible by official restrictions on gold circulation. This process was nearly finished, however, and unless a greater proportion of new gold went to monetary uses – which was unlikely unless it became more expensive (unless the price level fell) – increases in reserves would soon be reduced to or below production. Furthermore, since gold coinage had been a source of reserves, its shift to the vaults of central banks overstated the increase in reserves (Johnson 1995, p.117).

The situation of a representative country at the time of its return to the gold standard at the prewar par in 1925 may be expressed in terms of (3), with 1913 values indexed at unity,

(3.1) , where Pg = 1, y grew at an average rate of 3%, and the other values are from Tables 2 and 3. For given K =1 and rates of increase in y, adjustments of P under the renewed gold standard depend on Gm and H. Beginning with the former, we can approximate (8) by


where the rate of increase of monetary gold equals gold production, g, which depends on the price level in t-1 relative to the cost of gold production, P*. Estimates of a and b were obtained as follows, assuming a continuation of the gold-production technology existing just prior to 1914. Using Cassel’s estimates cited above, the years preceding 1914 were not far from a steady equilibrium, when annual rates of change in the gold stock and U.S. income were both near 3%, implying= a = 0.03. Then b = 0.6 follows from Pt-1 = 2.02 and = 0.02, as in the 1920s. This implies

(9) t = 1, 2, …where t = 1 is 1926.

Pt-1 = 2.02 implies 1% deflation. These data were predicted by Cassel (1920) and observed by Keynes (1929). Annual production less additions to the industrial arts and absorptions by the East left about 2% for the monetary stock. Given the “annual rate of increase of the world’s requirements due to the expansion of its economic life [of] about 3%,” as “generally estimated,” world prices will have to fall on the average by 1% per annum,” unless “central banks … economise in their gold habits.” We will see that they did just the opposite.

With few exceptions, policymakers and economists abstracted from the determinants of the gold stock in the 1920s. Senior had emphasized the effect of cost on production, although Mill (1848, p. 503) thought the lags were long. Wicksell (1898, pp.29-37) thought production was so small relative to the stock that it could be ignored. He was like most writers in taking the monetary gold stock as given. Figure 2 compares gold’s production and value (v = Pg/P) between 1835 and 2007, indexed to 1 in 1913. Cause and effect are difficult isolate, as is common in markets, because p and v sometimes move together and sometimes oppositely, as demand and production shocks alternate. Some movements are consistent with lagged technological and search responses to changes in v: the production increases in the 1850s, 1890s, 1930s, and 1980s followed increases in v; major declines occurred during and after wartime inflations.6 However, the most statistically significant relation between g and v occurred during 1910-39, when large changes in v (due to changes in both numerator and denominator) were determined by governments.

The left side of Table 4 shows estimates for this period. The right-hand side of the table shows, on the other hand, that production had little influence on the value of gold during this period. The failure of central bankers and most economists to take account of the influence of high (above prewar) prices on gold production is understandable in light of the imperfect relation before the war (although Figure 1 shows a fall in gold production during the suspension of the Napoleonic Wars). The failure of a similar response during the American Civil War may have been because the United States was alone in its suspension, while a free market for gold remained even there.

Table 4.

Regressions between dg and dv, 1910-39















































R2 / DW



Notes: Adjusted R2; Durbin-Watson statistics;

absolute t statistics in parentheses. Sources: See Figure 2.

In fact the relations operating during this period contained the seeds of a faster resumption than is suggested by these data. Cassel wrote in 1928 (pp.18-19):

In Denmark and Norway an attempt was made to restore the old gold parity by aid of a gradual and inappreciable process of deflation over an indefinite series of years. The experience of these countries forms the best proof of the impracticability of this policy. When once international speculation came to believe that a restoration of the old gold standard was to be expected, it took the currency into its own hands, and the authorities lost all control over developments. They were simply abliged to precipitate the deflation in order that the purchasing power of the currency should be made to correspond with its international value. In this way both currencies have been practically restored to their old gold parity, but the deflation which had to be gone through seriously affected the economic life of the countries. The losses were heavy, and unemployment became a most distressing factor.

Commercial bank deposits in Denmark and Norway fell 39% and 31%, respectively, between the ends of 1921 and 1926, comparable to the 31% fall in the United States between 1929 and 1932, and reminiscent of the market’s expedition of Ricardo’s resumption plan of 1819 (Wood 2005, pp.55-57).

Deflation was substantial even before the onset of the Great Depression. The median price fall of 24% between 1925 and 1928 exceeded that of 16% between 1928 and 1933. The United States was able to limit its price fall to 6% in the earlier period, but the one-third fall between 1925 and 1933 was comparable to the rest of the world. The fall in bank deposits reestablished the 1913 gold-reserve ratio: 12,005/4859 = 2.48 approximates the median 1933:1913 deposit ratio of 2.52.

Much has been made, at the time and since, of the maldistribution of reserves caused especially by France’s acquisition and America’s retention. This was alleged to have forced monetary contractions by low-reserve countries without offsetting expansions by countries determined to maintain reserves – who, however, suffered in their turn from the loss of foreign markets. It had been hoped that the gold standard could be maintained without deflation “by the aid of a systematic gold-economising policy aiming at such a restriction of the monetary demand for gold [as in the numerical example in the Appendix derived from the reduction in gold coin] as would prevent a rise in the value of that metal…. From 1928 onwards, however, this policy was completely frustrated by extraordinary demands for gold which brought about a rise in the value of gold of unparalleled violence” (Cassel 1932). Keynes wrote in early 1929:

From the days of the Genoa Conference in 1922 anxiety has often been expressed whether the world’s stock of gold would be adequate to its needs in the event of the great majority of countries returning to the gold standard. Professor Cassel has been foremost in predicting a scarcity. I confess that for my own part I did not, until recently, rate this risk very high. For I assumed – so far correctly – that a return to the gold standard would not mean the return of gold coins into the pockets of the public; so that monetary gold would be required in future solely for the purpose of meeting temporary adverse balances on international account….7 Accordingly – so I supposed, and here I was wrong – the monetary laws of the world would no longer insist on locking up most of the world’s gold as cover for note issues. For the contingency against which such laws had been intended traditionally to provide, namely, the public wishing to exchange their notes for gold, was a contingency which could no longer arise when gold coins no longer circulated. Moreover, to meet an adverse international balance, bill and deposits held at foreign centers would be just as good as gold, whilst having the advantage of earning interest between-times [the gold-reserve system].

But I was forgetting that gold is a fetish. I did not foresee that ritual observances would, therefore, be continued after they had lost their meaning. Recent events and particularly those of the last twelve months are proving Professor Cassel to have been right. A difficult, and even a dangerous, situation is developing ….

An important source of the demand for gold was recent British and French legal impositions of minimum reserve holdings [similar American legislation was older], which are “locked up and might just as well not exist for the purposes of day-to-day policy.”

Great emphasis has been placed on the demand for gold reserves as the immediate, and unnecessary (Bordo, Choudhri, and Schwartz 2002; Hsieh and Romer 2006), cause of the Great Depression. Several considerations, some already discussed, qualify or contradict this view. First, who is to say these demands were unnecessary?

Fetish or not, countries apparently wanted gold reserves similar to their prewar ratios. Whatever economists might regard as sufficient, the Fed was fettered if it thought it was. We might even wonder why, in light of the mispricing of gold, the desired ratios were not greater (Einzig 1935, pp.142-43). Furthermore, the uneven distribution of reserves was not unusual (Rist 1931). Prewar observers had marveled at the Bank of England’s operations on the basis of a “thin film of gold” (Sayers 1951), made possible by a commitment to the gold value of sterling.

Unfortunately the policies, or rather hopes, of the postwar central banks were bound to be disappointed. The former deputy-governor of the Bank of France wrote:

The idea that such an enormous loss of the buying power of gold, as resulted in the War and after the War, could be maintained, has always seemed to me chimerical. I have often discussed that question with … Benjamin Strong [of the Federal Reserve Bank of New York]. He was of quite another opinion. He thought that, with the big banks and the possibility of enormous credits, one would be able to maintain prices much better than one could do it before the war. I remember, however, his words during his last visit in Paris in 1928, [when] he admitted frankly: ‘Well, till now the facts have proved you right’ (Rist 1931).

“It is the first time in the history of financial doctrines,” Rist observed, “that the maldistribution of gold between the different countries has been put forward to explain the general movement of prices.” Finally, the rising price of gold presumably induced speculative demands by central banks as well as others.

Rist and Cassel are supported by Figure 3, which reinforces Figure 1 in showing the correspondence between price levels and exchange rates between 1913 and 1933. Those countries whose exchange rates approached their prewar pars (X1933/X1913 = X = 1.8 Those resuming the gold standard at devalued currencies (Italy, France, Belgium the three observations on the right) experienced corresponding price changes. These observation support the Eichengreen-Sachs (1985) argument that the “competitive” devaluations of the Great Depression were not at the expense of other countries, whose prices conformed to their own exchange rates.
4. Conclusion

The way in which the [League of Nations] Gold Delegation presents the causes of the breakdown of the gold standard seems to me entirely unacceptable. What we have to explain is essentially a monetary phenomenon, and the explanation must therefore be essentially of a monetary character. An enumeration of a series of economic disturbances and maladjustments which existed before 1929 is no explanation of the breakdown of the gold standard. In fact, in spite of existing economic difficulties, the world enjoyed up to 1929 a remarkable progress. What has to be cleared up is why the progress was suddenly interrupted.

Cassel, “Memorandum of dissent,” 1932.

Several writers have noted that economic performance in the 1920s compared favorably with pre-1914, and technological and other changes, while considerable, were not of a different order than earlier periods (Janssen, Mant, and Strakosch 1932; Aldcroft and Richardson 1969; Alford 1972). Furthermore, no one has been able combine these “causes” into an explanation why the post-1928 deflation was so great compared to others. The model presented here resolves the deflation portion of the quandary. The price fall was dictated by the normal operations of the gold standard. It did not “break down.”

Only a small minority recognized this at the time. One was the financial journalist Paul Einzig (1935,, who confessed that he had believed the French accumulation of reserves was largely responsible for the “crisis,” but came to the realization “that given the circumstances in which the currencies were stabilized between 1925 and 1929, a crisis of first-rate gravity was a mere question of time,” regardless of French monetary policy.9 For whatever reason, possibly because of the growing faith in the possibilities of official policies, even this minority was lost.


Appendix. The Money Multiplier

Let γ be central bank gold Gc relative to its liabilities (notes N); α be private currency (gold or notes, Gp,and Np) relative to commercial bank deposits D; ρ be commercial bank reserves (gold and notes, Gb,and Nb) relative to D; N = Np + Nb; exogenous Gm = Gc + Gp + Gb; and M = D + Gp + Np. Then M = (1 + α)Gm = HGm and

Gc = γ[ρ(1 - ρg) + α(1 - αg)]Gm/Δ, where Δ = γ(ρ + α) + (1 – γ)(ρρg + ααg), and ρg and αg are the gold proportions of deposits and currency.

For γ = 0.4, ρ = α = 0.1, and Gm = 1, (M, Gc) = (12.28, 0.82), (13.35, 0.93), and (13.75, 1) as ρg = αg = 0.08, 0.02, 0. That is, a reduction in the private demand for gold relative to bank deposits from 8% to 0 increases M by 12% (12.28 to 13.75).



Aldcroft, D.H. and H.W. Richardson. 1969. The British Economy 1870-1939. London: Macmillan.

Alford, B.W.E. 1972. Depression and Recovery? British Economic Growth 1918-39. London: Macmillan.

Altman, O.L. 1958. “A note on gold production and additions to international gold reserves,” IMF Staff Papers, April 258-88.

Bernanke, B. 1983. “Nonmonetary effects of the financial crisis in the propagation of the Great Depression,” AmerEconRev., 257-76.

_____ and K. Carey. 1986. “Nominal wage stickiness and aggregate supply in the Great Depression,” QuarJEcon, 853-83.

_____ and H. James. 1991. “The gold standard, deflation, and financial crisis in the Great Depression: An international comparison,” in R.G. Hubbard, ed., Financial Markets and Financial Crises. Chicago: Univ. Chicago Press.

Barro, R. 1979. “Money and the price level under the gold standard,” EconJ, March 13-33.

Bloomfield, A.I. 1959. Monetary Policy under the International Gold Standard. Federal Reserve Bank of New York.

Bordo, M.D. 2003. “Comment on ‘The Great Depression and the Friedman-Schwartz hypothesis’ by Christiano, Motto, and Rostagno,” J.Money, Credit, and Banking, Dec., pt. 2, 1199-1203.

_____, E.U. Choudhri., and A.J. Schwartz. 2002. “Was expansionary monetary policy feasible during the Great Contraction? An examination of the gold constraint,” Explor.inEcon.Hist., 1-28.

_____, C. Erceg, and C. Evans. 2000. “Money, sticky wages, and the Great Depression,” AmerEconRev, 1447-63

Carter, S.B., eds. 2006. Historical Statistics of the U.S.: Millennial Edition. New York: Cambridge University Press.

Cassel, G. 1920. “Further observations on the world’s monetary problem,” EconomicJ, March, 39-45.

_____. 1922. Money and Foreign Exchange after 1914. Macmillan.

_____. 1928. Postwar Monetary Stabilization. NY: Columbia Univ. Press.

_____. 1932. “Memorandum of dissent,” League of Nations (1932b, pp.74-75).

Chari, V.V., P.J. Kehoe, and E.R. McGrattan. 2002. “Accounting for the Great Depression,” AmerEconRev., May, 22-27.

Christiano, L., R. Motto, and M. Rostagno. 2003. “The Great Depression and the Friedman-Schwartz hypothesis,” J.Money, Credit, and Banking, Dec., pt. 2, 1199-1203.

Clarke, S.V.O. 1967. Central Bank Cooperation, 1924-31. Federal Reserve Bank of New York.

Cole, H. and L. Ohanian. 1999. “The Great Depression in the United States from a neoclassical perspective,” Federal Reserve Bank of Minneapolis QuarRev, 2-24.

1   2   3

Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur © 2016
rəhbərliyinə müraciət

    Ana səhifə